Emergence of a three‐sample conditional discrimination as foundation for reasoning capabilities
A simple 'same' symbol during training makes three-way conditional discriminations emerge for almost every learner.
01Research in Context
What this study did
The team taught adults to pick pictures that go together in groups of three.
They first showed pairs like A-B and B-C. Then they added a new cue X that meant 'same group'.
Some people got the X cue during training. Others did not. The researchers watched who could later do three-way matches without being taught.
What they found
Almost everyone who saw the X cue passed the new three-way test.
Only one person out of six passed when the cue was missing.
The cue acted like a green light telling the brain which items belong together.
How this fits with other research
Pérez-González et al. (2003) already showed that context symbols can guide untaught matches. The new study uses the same trick, but pushes it further to three-item sets.
Preston (1994) proved that after A-B and B-C training, people will spontaneously link A-C. Pérez-González et al. (2023) add the X cue so the leap from pairs to triplets becomes reliable instead of hit-or-miss.
McGee et al. (2024) tried a different boost: they told adults to 'picture the image' and also got strong emergence. Both papers agree that a small add-on—words or symbols—can double your success rate.
Why it matters
If you want clients to infer bigger sets from smaller taught units, drop in a clear context cue like a color or word that means 'same'.
It costs one extra stimulus per lesson and can turn sporadic emergence into a sure thing. Try it next time you teach categories, analogies or sorting tasks.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Add a small icon or color that means 'same group' while you teach A-B and B-C relations, then test untrained A-C-B triplets.
02At a glance
03Original abstract
We hypothesized that a three-sample conditional discrimination can emerge as a result of learning conditional discriminations with relational stimuli. After learning three first-order conditional discriminations AB, PQ, and CD, we taught a second-order conditional discrimination XAB in which X1 indicated selection of related stimuli (e.g., A1 and B1) and X2 of unrelated stimuli (e.g., A1 and B2). Then, we probed the emergence of conditional discriminations PQX and XCD in which the X stimuli were comparisons and contextual stimuli, respectively. Finally, a conditional discrimination was probed with stimuli P, Q, and C as samples and D1 and D2 as comparisons. When the P and Q stimuli were related (and related to X1 in PQX), all participants selected the D stimulus that was related to the C stimulus (D1 when C1 was present and D2 when C2 was present); when the P and Q stimuli were unrelated (and related to X2 in PQX), they selected the D stimulus unrelated to the C stimulus (D2 when C1 and D1 when C2), which demonstrated emergence based on the relations established among all stimuli. In Experiment 2, the teaching of XAB was omitted and only one in six participants demonstrated emergence, which indicated that relational stimuli X1 and X2 played an important role in emergence. Thus, a new type of emergence that mimics analogical reasoning was demonstrated. The obtained outcome suggests that this procedure provides a learning foundation for acquiring reasoning capabilities.
Journal of the Experimental Analysis of Behavior, 2023 · doi:10.1002/jeab.877